A MIMO (Multiple-Input Multiple-Output: multi-inputs and multi-outputs) transmission technology where a plurality of antennas are used for transmitting/receiving, and a plurality of different data series (data streams) are multiplexed spatially in the same frequency band, and concurrent communication is carried out has been put into practical use in a wireless LAN and a cellular system or the like. In a single user MIMO (Single User MIMO) where a plurality of different data series are spatially multiplexed and transmitted to certain one terminal apparatus, in order to enhance performances in separation and detection of the plurality of series of data series in the terminal apparatus, there is a method where transmission is carried out after precoding is applied to a transmission signal in a base station apparatus.
In addition, in a next generation cellular system or the like, proposed is a system where the number of transmission antennas with which a base station apparatus (transmission apparatus) are provided will be increased significantly as compared with the number of reception antennas with which a terminal apparatus (reception apparatus) are provided. In a system like this, in order to utilize effectively the transmission antennas of the base station apparatus and enhance further a system throughput, a multiuser MIMO (Multi-User MIMO) where data series addressed to a plurality of terminal apparatuses (users) are MIMO-multiplexed is proposed. However, since among a plurality of terminal apparatuses which receive the signals multiplexed by the multiuser MIMO, signals received by other terminals are not able to be known, performances will have deteriorated significantly in the present state due to multi-user interferences (Multi-User Interference: MUI) occurred among streams addressed to each user.
Here, when the base station apparatus knows CSI (Channel State Information) which is a channel status information from each transmission antenna of the base station apparatus to each reception antenna of each terminal apparatus, a transmission signal which is able to suppress MUI at the time of reception in the terminal apparatus without giving a large load to the terminal apparatus is able to be generated, and several methods for the purpose of that have been proposed (Non-Patent Literature 1). For example, in order that the reception may be carried out in a state where MUI is suppressed at the time of reception in the terminal apparatus, there is a method where transmission is carried out after a precoding is applied to the transmission signal in the base station apparatus.
As the example, there are Zero-forcing (ZF) precoding which from a channel matrix H which has, as an element, a complex channel gain calculated from the CSI between each transmission antenna and each reception antenna of each terminal apparatus, carries out weighting on a transmission signal (the transmission signal is multiplied by W=H−1) using an inverse matrix H−1 thereof (or pseudo-inverse matrix H†=HH(HHH)−1: a superscript H denotes a Hermite conjugate) as a weighting matrix (linear filter) W, and a linear precoding (beam forming) which carries out precoding of the transmission signal by linear processing, such as an MMSE precoding which carries out the weighting of the transmission signal base on the weighting matrix (linear filter) W=HH(HHH+αI)−1 (I denotes an identity matrix and α denotes a normalization coefficient) calculated by a minimum mean squared error (Minimum Mean Square Error: MMSE) norm.
In addition, as an example of another linear precoding, candidates (referred to as a code book) of a vector (constituent of weighting matrix) used for the linear precoding in the base station apparatuses are made to be determined in advance as known ones in transmitting/receiving, and a vector which will be able to be received in the most satisfactory performances in those candidates is made to be selected based on a channel estimation result in each terminal apparatus, and the information (Precoding Matrix Indicator: PMI) with respect to the selected vector is made to be fed back to the base station apparatus, and the base station apparatus multiplexes and transmits transmission data addressed to each terminal apparatus by the multiuser MIMO base on the linear precoding to a plurality of terminal apparatuses based on those fed back PMI (Non-Patent Literature 2).
As other examples, there is a nonlinear precoding which carries out precoding of the transmission signal based on nonlinear processing, where an interference signal component calculated from the CSI is made to be subtracted from the transmission signal in advance, and in order to suppress a transmit power which will have increased after the interference subtraction, the signal after the interference subtraction is made to be encoded at a signal point where the transmit power may be made to be reduced in a signal space. As one of this nonlinear precoding, Tomlinson-Harashima precoding (Tomlinson-Harashima Precoding: THP) which is capable of suppressing an increase in the transmit power based on applying a Modulo (Modulo, surplus) operation to signals in both of transmitting/receiving apparatuses has been proposed (Non-Patent Literature 3, Non-Patent Literature 4, Non-Patent Literature 5). At this time, also in the terminal apparatus which is a receiving side, the same Modulo operation as in the transmission with respect to the reception signal is made to be applied. Besides, the nonlinear process here mainly means one where processing in which a break point exists in outputs of the Modulo operation or the like is used.
In addition, based on the THP Modulo operation, in consideration of the fact that signal point arrangement at the time of a modulation in the base station apparatus is received by the terminal apparatus in a form repeated by a Modulo width, a method to calculate a logarithmic likelihood ratio (Log Likelihood Ratio: LLR) of demodulated bits (demodulation processing based on soft decision) from a Euclidean distance between a reception signal point (with noises added thereto) and each candidate signal point repeated by the Modulo width has been proposed (Non-Patent Literature 6).